Electric Strength of Solid Dielectrics. 



Table VIII.— Oiled Canvas * . 

 A=5-4.10- 6 . B = 9-62.10- 10 . A/B = 5720. 



129 





Ereakdown Voltage. 





Thickness. 





Diff. per 









Observed. 



Calculated. 



cent. 



•01 cm. 



1,500 



1,465 



-2-2 



•02 



2,500 



2,545 



+ 1-9 



03 



3,500 



3,440 



-16 



04 



4,250 



4,230 



-0-3 



•05 



4,900 



4,935 



+0-7 



•06 



5,500 



5,580 



+1-5 



•07 



6,000 



6,170 



4-2-9 



•08 



6,500 



6,730 



4-35 



•09 



6,900 



7,190 



4-4-1 



•10 



7,250 



7,790 



4-7*5 



The differences are in general within the range of experi- 

 mental error. At thicknesses much beyond working limits 

 the tv curves in some cases approximate to straight lines and 

 t — c-V Av is the limiting form j the c term indicates that there 

 is a stage which is completed. In equation (4) the quadratic 

 term is that of the first polarization, the linear term that of 

 interattraction. Since in t = c + Av the whole effect is linear, 

 the suggestion is that when this relation occurs interattraction 

 is so much greater in magnitude than the applied field that the 

 first polarization, though included, is quite masked. From 

 Table IX. porcelain and mica might be expected to approach 

 this condition. 



The above tables show that the relation t = Av + Bv' 2 is a 

 much closer approach to the observed values than any 

 formula of the shape t=cv 3 <' 2 can be, for in that case the co- 

 efficients A should be zero. Steinmetz's values of the break- 

 down strength of air at 100 alternations a second between disks 

 I find to fit closely the formula £ = 5'3 . 10" 5 r-f I'll . 10" 9 v 2 . 

 He remarks, " The tests made by Warren de la Rue, in which 

 a chloride of silver battery was employed, and which are there- 

 fore the only continuous potential tests free from the objection 

 due to the electrostatic machine, agree completely with the 

 formula t = av + bv* over the whole range up to 11,330 volts/' 



It may, therefore, be concluded from equations (3) and (4), 

 and the fact that the latter is found to give a close approxi- 

 mation to the experimental results, that the work done in 

 breaking down any solid dielectric is constant for that material, 

 and is independent of the thickness. It is possible that the 

 same law may hold for liquids and gases, though the expe- 

 rimental results for gases other than air are at present too 

 meagre for this to be tested. 



* S. P. Thompson, ' Dynamo Design,' fig. 23. 

 Phil. Mag. S. 6. Vol. 30. No. 175. July 1915. K 



